Conscious mental qualities, aka phenomenal qualities, are seemingly a leading factor in much of our behaviour. Pains make us recoil from painful stimuli, itches make us scratch, feelings of anger sometimes make us shout, visually perceiving red leads us to halt at stop lights, and so on. To relinquish this claim about the efficacy of conscious mental qualities would mean surrendering a major component of our everyday, intuitive self-conception; hence, the claim enjoys considerable prima facie plausibility. Unconscious mental qualities, however, (...) have been posited by a mere handful of philosophers historically, and are nowadays almost universally rejected. Via a case study of ‘restless legs syndrome’, I argue that there is a hitherto unnoticed cost to this prevailing rejection of unconscious mental qualities: the causal efficacy of conscious mental qualities is threatened. In fact, I argue, we face a dilemma: To endorse epiphenomenalism about conscious mental qualities, or to posit unconscious mental qualities. Since it is so plausible that conscious mental qualities are causally efficacious, this reasoning constitutes an argument for unconscious mental qualities. Moving beyond the sensory case, I explain how analogous reasoning might apply to other mental faculties where phenomenal qualities seem causally involved: notably, emotions and mental imagery. (shrink)